Vocative, where do you hang out in wh-interrogatives?

1 Introduction

This paper contributes to the literature on vocative constructions that co-occur with a sentence and their syntactic analysis. Focusing on Spanish, we explore vocatives that are expressions used to designate the hearer, and the impact that the semantic-pragmatic distinction of vocatives (i.e., calls and addressees: Zwicky 1974) has on the syntax.

The literature on vocatives distinguishes between those that constitute an independent utterance (1a), and others that are part of another utterance (1b). We focus on the latter one here.

Most related studies have analyzed vocative constructions in declaratives. In these clauses, three different positions for vocatives can be identified: initial, middle, and final, as shown in the following examples:

To the best of our knowledge, there is no research on vocatives co-occurring with wh-interrogatives. We believe that this combination provides interesting new insights into the syntactic analysis of vocatives. Before we examine possible vocative positions in wh-interrogatives, we describe some basic characteristics of vocatives and Spanish wh-interrogatives in Section 2 and Section 3. Sections 4 and 5 present the methodology and the results of our acceptability judgment task that we discuss in Section 6. In Section 7, we propose a new syntactic analysis within the cartographic framework that reflects the properties of both wh-interrogatives and vocatives. Finally, Section 8 draws some conclusions and highlights paths for future research.

2 Classification of vocatives

The term vocative comes from the Latin noun vocatīvus which is derived from the verb vocare. The literal meaning of vocare is ‘to invoke, call, name a person or thing personified’ (Brandimonte 2011: 251; our translation). A lot of research has emerged to define vocatives. The Real Academia Española (RAE) propose the following:

[C]onstituyen EXPRESIONES VOCATIVAS los nombres, los pronombres y los grupos nominales que se usan para llamar a las personas o animales (¡Eh, tú!; ¡Papá!, ¿me oís?; Lucera, ven acá), para iniciar un intercambio verbal o para dirigir a alguien un saludo (¡Hola, Clara!), una pregunta (¿Está cansado, don Marcelo?), una petición o una orden (Márchate, niña), una advertencia (Manuel, ten cuidado), una disculpa (Lo siento, caballero), etc. (RAE/ASALE 2009: §32.2g).^[1]

Therefore, we use the term vocative to describe nominal expressions that assign the addressee. Vocatives have been characterized from different perspectives: (i) morpho-syntactic perspective (i.e., class of words and phrases, positional mobility), (ii) semantic-pragmatic perspective (i.e., meaning, primordial functions, speaker-listener relationship), and (iii) phonological perspective (i.e., intonation). Concerning the morpho-syntactic perspective, vocatives can appear in three different positions (see [3]), namely: initial, middle, and final (Espinal 2013; Hill 2007, 2013b; Stavrou 2014; Zwicky 1974). These three positions are the topic of our paper.

Regarding the semantic-pragmatic perspective, vocatives have different functions depending on their positions. According to Zwicky (1974), initial vocatives (3a) are used to attract the attention of the interlocutor (call function), while middle (3b) and final vocatives (3c) maintain the contact between speaker and listener (addressee function). Consequently, middle and final vocatives have the same basic semantic-pragmatic function. However, they differ with respect to the context in which they are used. Middle vocatives are used to express politeness and occur therefore more often in formal speech, while final vocatives are more frequent in informal speech, as shown by Edeso Natalías (2005) and Kleinknecht (2019).

With respect to the phonological perspective, vocatives have their own intonation, regardless of their position or their semantic-pragmatic function. They are interpreted as independent intonation phrases (I). Put differently, their intonation is independent from other constituents or intonational phrases of the utterance (U) (Prieto and Roseano 2010). Their phonic properties are also essential to distinguish them from other elements, such as subjects. In contrast to vocatives, the intonation of the subject is determined by the utterance (Prieto and Roseano 2010):

3 Word order in wh-interrogatives

In contrast to Spanish declaratives, the word order in wh-interrogatives is fairly restricted (e.g., Francom 2012). The wh-phrase is fronted to the sentence-initial position. This, in turn, triggers the reordering of the verb and an overt subject if it is not identical to the wh-phrase. This phenomenon is known as subject-verb inversion, or inversion, and occurs obligatorily in wh-interrogatives (5a). Therefore, preverbal subjects lead to ungrammaticality (5b) (Barbosa 2001; Dumitrescu 2016; Torrego 1984).

However, subject-verb inversion is not always obligatory in wh-interrogatives. For Peninsular Spanish, there are at least two cases in which subjects can occupy the preverbal and postverbal position. The first exception concerns causal wh-phrases, such as por qué (‘why’), as already noticed by Torrego (1984) and Suñer (1994):

Empirical work on this topic confirms this finding and shows that por qué (‘why’) is highly accepted with both preverbal and postverbal subjects (Adli 2010; Goodall 2004; Schmid et al. 2021):

The second exception concerns complex or so-called d-linked wh-phrases (Contreras 1999; Fabregas and Mendívil 2013; Francom 2012; Goodall 2005; Martín 2003). D-linked wh-phrases refer to a set of referents that are familiar from the discourse context. They also allow subjects to occur in the preverbal position:

In addition to preverbal subjects that can occur between the wh-phrase and the verb, other constituents, such as complements (9a), can also intervene between the wh-phrase and the verb in wh-interrogatives introduced by por qué (‘why’) or d-linked wh-phrases (Goodall 1993; Martín 2003; Suñer 1994). This does not apply to interrogatives with argumental wh-phrases (9b).^[2]

In sum, related literature has shown that in contrast to bare wh-interrogatives, interrogatives introduced by ‘why’ and d-linked wh-elements allow subjects and complements to intervene between the wh-element and the verb, suggesting that the same difference can be observed for middle vocatives. To date, no study has investigated whether vocatives underly the same restrictions as subjects in wh-interrogatives or if they can appear in the initial, middle, and final position, as in declaratives (see [3]). We will fill this gap by answering the following research questions (RQs):

1. In which position are vocatives accepted in wh-interrogatives? (RQ1)

2. Do d-linked wh-phrases and por qué (‘why’) pose less restriction on the middle positions of vocatives than bare wh-elements? (RQ2)

3. Which position(s) do initial, middle, and final vocatives occupy in the left periphery? (RQ3)

4 Methodology

We conducted a web-based acceptability judgment task to investigate the positions in which vocatives are accepted and whether wh-phrases like por qué (‘why’) and d-linked wh-phrases pose less restriction on the position of vocatives than bare wh-elements (RQ1 and RQ2).

5 Results

Figure 1 summarizes the results of the acceptability judgment task and plots the mean ratings of the three vocative positions (initial, middle, and final) across the wh-phrases (por qué [‘why’], d-linked wh-phrases and qué [‘what’]). The results suggest that participants, on average, show the highest ratings for the initial vocative position, followed by ratings for the final vocative position. We also observe that these two positions receive similar ratings for all three wh-phrases. The middle vocative position, overall, receives the lowest ratings but shows a remarkably great difference in ratings across the wh-phrases. For por qué (‘why’) and d-linked wh-phrases, participants have higher ratings, and they seem to accept (at least marginally) the middle vocative position. By contrast, for interrogatives with qué (‘what’), this position shows extremely low ratings.

Figure 1:

Mean ratings for vocative positions (initial, middle, and final) across wh-phrases with 95% confidence intervals.

We ran a generalized additive mixed model to test which vocative positions participants accept in wh-interrogatives (RQ1) and whether d-linked wh-phrases and por qué (‘why’) pose less restriction on the middle positions of vocatives than bare wh-elements (RQ2). Due to a significant interaction between vocative position and wh-phrase, we could not assess the effect of the two variables and thus calculated a separate generalized additive mixed model for vocative position in combination with each wh-phrase and for wh-phrase in combination with each vocative position. For initial vocatives, we did not observe a difference between the three wh-phrases: wh-interrogatives with d-linked wh-phrases (β = 0.12, SE = 0.21, z = 0.60, p = 0.55) and qué (‘what’) (β = 0.22, SE = 0.21, z = 1.10, p = 0.27) did not differ significantly from those introduced by por qué (‘why’).

The ratings for final vocatives were significantly lower for all wh-phrases (d-linked: β = −1.04, SE = 0.19, z = −5.42, p < 0.001; por qué (‘why’): β = −0.83, SE = 0.19, z = −4.39, p < 0.001; qué (‘what’): β = −0.80, SE = 0.20, z = −4.07, p < 0.001), but did not significantly differ between wh-phrases. Neither wh-interrogatives introduced by d-linked wh-phrases (β = −0.08, SE = 0.17, z = −0.44, p = 0.66) nor by qué (‘what’) (β = 0.26, SE = 0.18, z = 1.43, p = 0.15) deviated significantly from those with por qué (‘why’). This implies that the differences were driven by vocative position and not by wh-phrase.

In the last step, we focused on the middle vocative position. This position also differs significantly from the initial position for all wh-phrases (d-linked: β = −2.83, SE = 0.19, z = −15.00, p < 0.001; por qué (‘why’): β = −3.07, SE = 0.18, z = −16.72, p < 0.001; qué (‘what’): β = −5.60, SE = 0.22, z = −25.73, p < 0.001). For middle vocatives, we also found an effect of wh-phrase. We observe significantly higher ratings for wh-interrogatives with d-linked wh-phrases than for wh-interrogatives with por qué (‘why’) (β = 0.44, SE = 0.16, z = 2.63, p < 0.001). For wh-interrogatives introduced by qué (‘what’), we observe significantly lower ratings (β = −2.21, SE = 0.20, z = −11.20, p < 0.01). This suggests that the differences were not only driven by the vocative position, but also by the wh-phrase, especially by qué (‘what’) (see Figure 1).

6 Discussion

The first research question dealt with the positions that vocatives occupy in wh-interrogatives (RQ1). For the initial and final vocative position, we found a high level of acceptability over all wh-elements. Although the ratings for the final vocative position were significantly lower than for the initial vocatives, the ratings differed only slightly between the two positions. These results indicate that Spanish speakers fully accept vocatives in the initial position, directly before the wh-phrase, as well as in the final position if they occur remotely from the wh-element at the right end of the clause. These results are consistent with literature on vocatives in declaratives, showing that the availability of these two positions in combination with declaratives depends on the intention of the interlocutor, and consequently on the semantic-pragmatic properties of the vocative, as outlined in Section 2.

For a closer analysis of the middle vocative position in wh-interrogatives, i.e., vocatives directly to the right of the wh-phrase, we turn to RQ2. This research question dealt with whether d-linked wh-phrases and por qué (‘why’) pose less restriction on the middle position of vocatives than other wh-elements like qué (‘what’). On average, we found the lowest ratings for middle vocatives in combination with all wh-elements. However, this does not mean that they are unacceptable at all. For por qué (‘why’) and d-linked wh-phrases, middle vocatives obtained mean ratings between 3 and 4, i.e., they seem to be accepted (at least marginally). These comparably lower ratings for middle vocatives are consistent with the results of Kleinknecht’s (2019) corpus study which show an increasing use of middle vocatives in formal speech. However, our items only contain informal situations for which final vocatives are more frequent. In interrogatives with qué (‘what’), middle vocatives received a mean rating of about 1. The violin plot in Figure 2 further confirms these findings since it shows the distribution of the ratings for the middle vocatives in wh-interrogatives introduced by por qué (‘why’), d-linked wh-phrases, and qué (‘what’). For the wh-element qué (‘what’), we did not find much variation. The shape of the graph reveals that besides a few outliers, participants did not accept middle vocatives in this context. However, for por qué (‘why’) and d-linked wh-phrases the silhouette of the plots shows that at least some participants highly accepted middle vocatives.

Figure 2:

Distribution of the ratings for the middle vocatives in wh-interrogatives introduced by por qué (‘why’), d-linked wh-phrases, and qué (‘what’).

Taken together, the results for interrogatives with por qué (‘why’) and d-linked wh-phrases clearly differ from those of qué (‘what’).^[4] We can conclude that middle vocatives are only ungrammatical in wh-interrogatives introduced by qué (‘what’) and other bare wh-elements, while they are (at least marginally) acceptable in interrogatives introduced by por qué (‘why’) and d-linked wh-phrases:

A similar behavior as for middle vocatives was observed by Cognola and Cruschina (2021) for the use of the particle poi ‘lit. then’ (as a discourse marker) in Italian. Poi in middle position is restricted to interrogatives with perché ‘why’ and is excluded in other wh-interrogatives. For Spanish, we assume the same distribution of the particles pues and entonces (lit. ‘then’) in middle position as we found for middle vocatives. While they can intervene between por qué (‘why’)/d-linked wh-phrases and the verb, they are excluded in middle position in wh-interrogatives with qué (‘what’):^[5]

In sum, our results indicate that the occurrence of middle vocatives in interrogatives introduced by d-linked wh-phrases and por qué (‘why’) is less restricted than in interrogatives with other wh-elements like qué (‘what’). In the former group, vocatives can intervene between the wh-phrase and verb, like subjects do, as shown in Section 3. In contrast, interrogatives introduced by qué (‘what’) or other bare wh-phrases do not allow any of these elements to intervene between the wh-word and the verb.

7 Syntactic analysis

7.2 Our analysis

The results presented in Section 5 as well as previous work on vocatives and wh-interrogatives raise the question of the position(s) initial, middle, and final vocatives occupy in the left periphery (RQ3). Since our results particularly provide new insights for middle and final vocatives, we organize this section as follows.

First, in Section 7.2.1, we relate our results on initial vocatives to previous findings of vocatives and wh-interrogatives. Second, in Section 7.2.2, we focus on middle vocatives and account for the asymmetry between wh-interrogatives introduced by bare wh-elements on the one side, and wh-interrogatives introduced by ‘why’ and d-linked wh-elements on the other. Finally, in Section 7.2.3, we consider the position of final vocatives in the left periphery. As shown in the previous section, related studies focus on the properties of initial and middle vocatives. Therefore, the position of final vocatives is usually linked to middle vocatives. It is true that an analysis that includes only one single Voc_addrP for middle and final vocatives is appropriate with respect to the economical aspect of language. Nevertheless, the results of our experiment cast doubt over the assumption of one single Voc_addrP for both middle and final vocatives. For our line of argument in this chapter, we include the results from our acceptability judgment task as well as other examples that are attested by different speakers of Peninsular Spanish.

7.2.1 Initial vocatives in the left periphery

As we have pointed out in Section 7.1, related literature agrees that the initial vocatives (i.e., calls) are located above ForceP:

(18)

[Voc(call)P Voc call ° [ForceP Forceº [TopP* Topº* [FocP Focº [TopP* Topº* [FinP Finº]]]]]]

The results of our acceptability judgment task confirm this analysis since initial vocatives are compatible with all kinds of wh-elements, namely bare wh-elements such as qué (‘what’), por qué (‘why’), and d-linked wh-phrases. Although these three types of wh-phrases occupy different positions in the left periphery, all are located below Voc_callP. First, qué and other bare wh-elements are moved to the left periphery (Goodall 1993; López 2013; Rizzi 1996; Suñer 1994), more precisely to [Spec, Foc] within the cartographic framework (Rizzi 1997, 2006), to check the [+WH] feature (López 2009):

(19)

[ForceP [TopP [IntP [TopP [FocP qué [TopP [FinP …]]]]]]]

Rizzi justifies this assumption by showing the incompatibility of bare wh-phrases with other focal elements (see [20]). Put differently, a bare wh-phrase and a fronted contrastive focus are in complementary distribution.

(20)

*¿Qué	a	Mario	le	deberíamos	decir	(,	y	no	a	Juan)?
what	prep	Mario	him	should-1pl	say		and	not	prep	Juan
‘What should we say to Mario and not to Juan?’

Rizzi’s (1997) assumption on the position of bare wh-phrases is compatible with previous analyses of initial vocatives (Espinal 2013; González López 2019; Moro 2003). We assume González López’s (2019) analysis for initial vocative and suggest that the Voc_callP occurs above these wh-phrases:

(21)

a.

Antonio,	¿qué	compraste	en	el	supermercado?
Antonio voc	what	bought-2SG	at	the	supermarket
‘Antonio, what did you buy at the supermarket?’

b.

[Voc(call)P Antonio [ForceP [TopP [IntP [TopP [FocP qué [TopP [FinP …]]]]]]]]

Second, por qué (‘why’) is proposed to be directly merged in IntP in the left periphery (e.g., Rizzi 1996, 2001; Stepanov and Tsai 2008; Thornton 2008) or in a position very close to the articulated CP (Shlonsky and Soare 2011). In Section 3, we provided first evidence for this claim by showing that por qué (‘why’) allows preverbal subjects and complements to intervene between the wh-phrase and the verb, while the other bare wh-phrases do not:

(22)

a.

¿Por qué	Juan	compró	un	coche	nuevo?
why	Juan	bought	a	car	new
‘Why did Juan buy a new car?’

b.

*¿Qué	Juan	compró	en	el	supermercado?
what	Juan	bought	in	the	supermarket
‘What did Juan buy in the supermarket?’

Another argument in favor of a distinct and higher position of ‘why’ comes from the fact that ‘why’ can co-occur with focus in a fixed order. The focal element can only follow the wh-phrase but cannot precede it (Rizzi 2001: 294):

(23)

a.

Perché	QUESTO	avremmo	dovuto	dirgli,	non	qualcos’altro?
why	this	have-1pl	should	say-him	not	something-else
‘Why should we have said this to him, not something else?’

b.

*QUESTO	perché	avremmo	dovuto	dirgli,	non	qualcos’altro?
this	why	have-1pl	should	say-him	not	something-else
‘This why should we have said this to him, not something else?’

Such focal elements interact differently with ‘why’-interrogatives than with bare wh-interrogatives (compare [24] and [25]). In ‘why’-interrogatives, the shift of focus to another constituent requires a different answer, while this is not the case for other wh-interrogatives. The questions in (25) can be answered in the same way, regardless of the focus value.

(24)

a.

Why did Adam eat the APPLE? – Because it (the apple) was the only food around.

b.

Why did Adam EAT the apple? – Because he couldn’t think of anything else to do with it.

(25)

a.

When did Adam eat the APPLE? – At 4 p.m. on July 7.

b.

When did Adam EAT the apple? – At 4 p.m. on July 7.

The analysis of a separate position for ‘why’ (Rizzi 2001) is compatible with the analysis on initial vocatives presented in (18), as can be seen in the following example:

(26)

a.

Antonio,	¿por qué	has	comprado	el	champú	más	caro?
Antonio voc	why	have-2SG	bought	the	shampoo	most	expensive
‘Antonio, why did you buy the most expensive shampoo?’

b.

[Voc(call)P Antonio [ForceP [TopP [IntP por qué [TopP [FocP [TopP [FinP …]]]]]]]]

Third, as we have shown in Section 3, d-linked wh-phrases also differ from other bare wh-elements and allow preverbal subjects to intervene between the wh-phrase and the verb. To account for this exception, Rizzi (2006) argues that d-linked wh-phrases have a different landing site from other bare wh-elements. More specifically, he assumes that d-linked wh-phrases target a topic position because they can be characterized in the same way as topics (Grewendorf 2002). Both d-linked wh-phrases and topics refer to a referent that is already introduced in the discourse content. However, a simple topic position cannot attract a wh-phrase. Therefore, Rizzi (2006) supposes that this topic position is also endowed with a Q feature as IntQ moves to the higher Top and creates a complex head. This movement converts the topic position into a landing site for d-linked wh-phrases, as shown in the following representation:^[9]

(27)

[ForceP [TopP d-linked wh-phrases IntQ+Top [IntP tInt … [FinP …]]]]

If Rizzi’s (2006) analysis is on the right track, d-linked wh-phrases and previous analyses for initial vocatives are compatible:

(28)

a.

Antonio,	¿de	qué	marca	has	comprado	el	champú?
Antonio voc	from	what	brand	have-2SG	bought	the	shampoo
‘What brand did you buy the shampoo from?’

b.

[Voc(call)P Antonio [ForceP [TopP de qué marca [IntP [TopP [FocP [TopP [FinP …]]]]]]]]

To sum up, the syntactic account for initial vocatives is compatible with previous analyses for all three types of wh-phrases regardless of the position they occupy in the left periphery. However, these different positions become crucial for the analysis of middle vocatives, as will be shown in the following section.

7.2.2 An approach to middle vocatives in the left periphery

The results of our acceptability judgment task have shown an asymmetry between wh-interrogatives introduced by bare wh-elements and wh-interrogatives with ‘why’ and d-linked wh-phrases. While middle vocatives are not acceptable in wh-interrogatives with bare wh-elements, they are acceptable (at least marginally) in those introduced by ‘why’ and d-linked wh-phrases. In this section, we present an analysis based on Slocum (2016) and Rizzi (2001, 2006 that accounts for this asymmetry.

Remember that González López (2019) and Slocum (2016) propose an analysis for middle vocatives in declaratives, as shown in (29). They suggest a different position, namely Voc_addrP, which differs from the position of initial vocatives due to their diverging semantic-pragmatic characteristics. According to them, the Voc_addrP is located between the FocP and the highest TopP, as shown in (29b).

(29)

a.

Yo,	Jessica,	quiero	ir	a	ver	una	película.
I	Jessica voc	want-1sg	go	to	see	a	film
‘I, Jessica, want to go to a movie.’
(Slocum 2016: 108, our translation)

b.

[ForceP Forceº [TopP* Topº* [ Voc(addr)P Voc addr ° [FocP Focº [TopP* Topº* [FinP Finº]]]]]

This analysis allows for three predictions concerning the grammaticality of middle vocatives in wh-interrogatives. First, wh-interrogatives with bare wh-elements do not allow middle vocatives since bare wh-elements are located below the Voc_addrP: in FocP, and the T-to-C movement that carries the wh-feature from T to the left periphery ensures that no other constituent can intervene between the wh-phrase and the verb (Rizzi 1997). The results of our acceptability judgment task show that middle vocatives are indeed unacceptable in wh-interrogatives with qué (‘what’) (30a). Thus, Slocum’s (2016) analysis accounts for this fact, as illustrated in (30b).

(30)

a.

*¿Qué,	Natalia,	has	señalado	en	el	texto?
what	Natalia voc	have-2sg	marked	in	the	text
‘What did you mark in the text, Natalia?’

b.

[ForceP [TopP* [ Voc(addr)P Natalia [FocP qué [TopP* [FinP …]]]]]]

Second, the co-occurrence of fronted foci and middle vocatives should also be excluded according to Slocum’s (2016) analysis due to similar reasons as for bare wh-interrogatives: Focus fronting does not allow another constituent to intervene between the fronted focus and the verb. For instance, it requires the reordering of subject and verb (Leonetti 2018). Example (31) confirms this prediction and shows that middle vocatives cannot be inserted between the fronted focus and the verb. Recall, a similar behavior has been observed by Ashdowne (2002) for cleft sentences, as shown in (15).

(31)

*El	bocadilloFoc,	María,	he	comido	(y	no	la	ensalada).
the	sandwich	María voc	have-1sg	eaten	(and	not	the	salad)
‘I have eaten THE SANTDWICH (and not the salad).’

Third, if Slocum’s (2016) analysis is correct, we expect that d-linked wh-phrases can co-occur with middle vocatives (see [32a]) because the highest TopP is their final landing side which is already endowed with the wh-feature (see Rizzi 2006 for details). This TopP is located above the Voc_AddrP and the wh-phrases can directly precede the middle vocative. Our experimental results reveal that Slocum’s (2016) analysis correctly predicts the occurrence of middle vocatives in wh-interrogatives introduced by d-linked wh-phrases, as shown in (32b).

(32)

a.

¿Con	qué	color,	Natalia,	has	señalado	la	palabra	“correr”?
with	what	color	Natalia voc	have-2sg	marked	the	word	run
‘With what color, Natalia, did you mark the word “run”?’

b.

[ForceP [TopP* Con qué color [ Voc(addr)P Natalia [FocP [TopP* [FinP …]]]]]]

Slocum’s (2016) analysis alone is not able to predict the position of por qué (‘why’) relative to the middle vocative in the left periphery, i.e., if middle vocatives can occur in wh-interrogatives introduced by por qué (‘why’). Remember, the literature on ‘why’-interrogatives has shown that ‘why’ allows different elements to precede the verb suggesting that it is also more permissive with respect to middle vocatives. The results of our experiment confirm this hypothesis and show that middle vocatives are accepted to a certain extent. This finding implies that ‘why’ is located higher in the syntactic structure than middle vocatives. The combination of Slocum’s (2016) and Rizzi’s (2001) analysis (see [33]) could explain this order if we assume that IntP is located above Voc_addrP, as shown in (34).

(33)

[ForceP [TopP* [IntP [TopP* [FocP [TopP* [FinP … ]]]]]

(Rizzi 2001: 289)

(34)

a.

¿Por qué,	Natalia,	has	señalado	este	párrafo?
why	Natalia voc	have-2sg	marked	this	paragraph
‘Why, Natalia, did you mark this paragraph?’

b.

[ForceP [TopP* [IntP por qué [TopP* [ Voc(addr)P Natalia [FocP [TopP* [FinP …]]]]]]

This assumption is reinforced by the placement of topics relative to the middle vocative. The middle vocative can be preceded and followed by a topic:

(35)

a.

¿Por qué	en	ese	lugar,	María,	todo	va	bien?
why	in	this	place	María voc	everything	goes	fine
‘Why is everything fine in this place, Maria?’

b.

¿Por qué,	María,	en	ese	lugar	todo	va	bien?
why	María voc	in	this	place	everything	goes	fine
‘Why is everything fine in this place, Maria?’

This variation of the topic position is also found in wh-interrogatives introduced by d-linked wh-phrases (36). This indicates that Voc_addrP is located between the second TopP and the FocP. Therefore, we assume the structure in (37) for the left periphery.

(36)

a.

¿A	qué	chica	en	ese	lugar,	María,	has	visto?
prep	what	girl	in	this	place	María voc	have-2SG	seen
‘Which girl have you seen in this place, María?’

b.

¿A	qué	chica,	María,	en	ese	lugar	has	visto?
prep	what	girl	María voc	in	this	place	have-2SG	seen
‘Which girl have you seen in this place, María?’

(37)

[Voc(call)P [ForceP [TopP [IntP [TopP [ Voc(addr) [FocP [TopP [FinP …]]]]]]]]]

7.2.3 An approach to final vocatives in the left periphery

For final vocatives, the related literature proposes different syntactic analyses (see Section 7.1). These analyses are mainly based on data with initial and middle vocatives, whereas final vocatives are treated only incidentally. Remember, Slocum (2016) and González López (2019) suppose a single Voc_addrP for middle and final vocatives. Their core argument for this single Voc_addrP to the left of the FocP is that addressees can only appear to the left of the focal constituent, as described in Section 7.1.

We argue that this analysis is problematic, especially in combination with different wh-interrogatives, and that addressees can indeed appear to the right of the FocP. We illustrate this issue with two examples. First, wh-interrogatives with bare wh-phrases, such as qué, can co-occur with final vocatives (38), as shown by the results of our acceptability judgment task. This result is unexpected in Slocum’s (2016) analysis because wh-elements like qué are moved to the FocP in the left periphery and freeze in this position (Rizzi 1997), so they cannot further be moved above the addressee.

(38)

¿QuéFoc	has	señalado	en	el	texto,	Natalia?
what	have-2sg	marked	in	the	text	Natalia voc
‘What did you mark in the text, Natalia?’

The second piece of evidence against Slocum’s (2016) analysis comes from wh-interrogatives with ‘why’. In these sentences, addressees can precede and follow focal material. The following example shows a focal object in both positions: to the right (see [39a]) and to the left of an addressee (see [39b]). Recall, according to Slocum (2016), the structure in (39b) should be excluded due to the characteristic of middle addressees to lie at the boundary between focus and background and her assumption that middle and final vocatives occupy the same position.

(39)

a.

¿Por qué,	María,	has	comido	el	bocadillo Foc
why	María voc	have-2sg	eaten	the	sandwich
(y	no	la	ensalada)?
(and	not	the	salad)
‘Why did you eat the sandwich (and not the salad), María?’

b.

¿Por qué	has	comido	el	bocadillo Foc	(y	no	la
why	have-2sg	eaten	the	sandwich	(and	not	the
ensalada),	María?
salad)	María voc
‘Why did you eat the sandwich (and not the salad), María?’

Note that this observation is not language-specific for Spanish or even a property of Romance languages. We find final vocatives in ‘why’-interrogatives with a focal object in other languages, such as German (40a) or Italian (40b). Consequently, there are many examples from different languages that remain unexplained.

(40)

a.

Warum	hast	du	das	Brot Foc	gegessen,	Maria?
why	have	you	the	sandwich	eaten	Maria voc
‘Why did you eat the sandwich, Maria?’

b.

Perché	hai	mangiato	il	paninoFoc,	Maria?
why	have-2sg	eaten	the	sandwich	Maria voc
‘Why did you eat the sandwich, Maria?’

To address these outstanding examples, a modified approach for final vocatives is required. We assume that there is a second Voc_addrP in addition to the one proposed by Slocum (2016) and González López (2019): an upper Voc_addrP above FocP, as described in Section 7.2.2, and a lower one below the FocP.^[10] We suggest that this lower Voc_addrP is located between the lowest TopP and the FinP, as shown in (41). The reason for this position is a change in information structure that we observe in the case of final vocatives: the information that was previously interpreted as new (see [42a]), is interpreted as given and especially highlighted, in (42b).

(41)

[ForceP [TopP [IntP [TopP [ Voc(addr)P [FocP [TopP [ Voc(addr)P [FinP …]]]]]]]]]]

(42)

a.

Hace	frío.
makes	cold
‘It’s cold.’

b.

Hace	frío,	Natalia.
makes	cold	Nataliavoc
‘It’s cold, Natalia.’

Similar to the declarative sentence in (42b), has comido [el bocadillo] _focus in the ‘why’-interrogative (39b) has to be given in the previous context and receives the interpretation of a reproach. The whole IP is the entity about which the reason asked for and can therefore be considered as a topic. Consequently, we assume that the IP underlies remnant topicalization (Belletti 2004) and the final vocative remains in-situ due to its special intonation, as proposed for initial and middle vocatives, as shown in Section 2 (for a similar analysis for Italian final particles, see Munaro and Poletto 2005, 2009). The syntactic analysis for (39b) is shown in the following example:

(43)

Consequently, our proposal allows for a uniform analysis of final vocatives (44) that are highly acceptable in all wh-interrogatives according to the results of our acceptability judgment task.

(44)

[ForceP [TopP d-linked wh-phrasesk [IntP why [TopP [Voc(addr)P [FocP whatj [TopP […]i [ Voc(addr)P Natalia [FinP … ti tj/tk]]]]]]]]]

This analysis further addresses the fact that Spanish allows for multiple addressees within one sentence (in contrast to English),^[11] as shown in (45a) for wh-interrogatives and (45b) for declaratives.

(45)

a.

¿Por qué,	Antonio,	has	regalado	este	libro	a	tu
why	María voc	have-2sg	given	this	book	to	your
hermano,	hombre?
brother	man voc
‘Why, María, did you give this book to your brother, man?’

b.

A	tus	amigas,	Ana,	las	vi	ayer,	cielo.
prep	your	friends	Ana voc	them	saw-1sg	yesterday	honey voc
‘Your friends, Ana, I saw them yesterday, honey.’

In the case of multiple addressees, we observe the same change in information structure, as described above. To illustrate this change, we reconsider example (16):

(46)

Jessica:	Creo	que	deberíamos	quedarnos	en	casa
Jessica	think-1sg	that	should-1pl	stay-ourselves	at	home
esta	noche.
this	night
‘Jessica: I think we should stay in tonight.’
Paul:	Yo,	Jessica,	quiero	ir	al	cine.
Paul:	I	Jessica voc	want-1sg	go	to-the	cinema
‘Paul: I, Jessica, want to go to a movie.’

Remember, in this example, the middle vocative marks the edge of the focus domain which is the right part of the sentence. However, if we add a final vocative to Paul’s sentence (see [47]), the information structure of the final part changes. It is no longer considered as new information, instead it is interpreted as given information that Paul reminds her of. Thus, quiero ir al cine receives a topic interpretation.^[12] The information structure of the subject Yo does not change and is still considered a contrastive topic.

(47)

Jessica:	Creo	que	deberíamos	quedarnos	en
Jessica	think-1sg	that	should-1pl	stay-ourselves	at
casa	esta	noche.
home	this	night
‘Jessica: I think we should stay in tonight.’
Paul:	Yo,	Jessica,	quiero	ir	al	cine,	cielo.
Paul:	I	Jessica voc	want-1sg	go	to-the	cinema	honey voc
‘Paul: I, Jessica, want to go to a movie, honey.’

The analysis in (48) reflects the information structure of example (47). The subject Yo moves to the highest TopP and quiero ir al cine undergoes phrasal movement to the specifier of the lowest TopP leaving the middle addressee in between. The final addressee follows the topicalized phrase.

(48)

In sum, we argue that the analysis of vocatives co-occurring with a sentence has to be revisited and that there are three positions for vocatives in the left periphery: the Voc_callP above the ForceP, where initial vocatives are located, and two Voc_addrP in the lower part of the left periphery for middle and final vocatives. (49) Provides an example of the left periphery including the three projections and showing the corresponding hierarchy.

(49)

[ Voc(call)P [ForceP [TopP [IntP [TopP [ Voc(addr)P [FocP [TopP [ Voc(addr)P [FinP ]]]]]]]]]]

7.2.4 The position of vocatives and subjects in wh-interrogatives: a comparison

Although vocatives and (preverbal) subjects are distributed equally in the three types of wh-interrogatives, their syntactic analysis reveals crucial differences. As shown above, ‘why’-interrogatives allow subjects and vocatives to intervene between the wh-phrase and the verb. Despite this parallelism, middle vocatives and preverbal subjects occupy different positions in the syntactic structure. For Italian ‘why’-interrogatives, Bianchi et al. (2017) show that the distribution of subjects depends on information structure (see Leonetti 2018 for a similar line of argumentation for Spanish). ‘Neutral’ (non-focal) subjects occur in preverbal position and are moved from their base position to a higher position in the syntactic structure. The related literature provides different accounts for the final landing sites of preverbal subjects. For example, Cardinaletti (2004) assumes that strong and overt preverbal subjects move to the Subj(ect)P(hrase) and weak or null subjects pronouns move to the Specifier of TP, while Barbosa (1995) and Alexiadou and Anagnostopoulou (1998) suggest that preverbal subjects have the same characteristics as clitic left-dislocated topics.^[13] Irrespective of which of these positions the preverbal subject moves to, middle vocatives are merged in the left periphery and occupy a different position, namely the VocaddrP, as shown in Section 7.2.2. In contrast to ‘why’-interrogatives, in bare wh-interrogatives, both preverbal subjects and middle vocatives are excluded. The wh-phrase is adjacent to the verb and the potential intervening subject or vocative has to be displaced. Put differently, wh-movement inhibits that first, the subject moves to a higher position and second, the occurrence of middle vocatives.

Not only the preverbal subject position but also the postverbal subject position differs from the position of vocatives. In Italian ‘why’-interrogatives, focal subjects typically occur in postverbal position and are presumably analyzed in FocP of the lower IP internal periphery (Belletti 2004). However, the assumption that this analysis fully applies to Spanish is problematic because Schmid et al. (2021) show that not only focal subjects, but also non-focal subjects are highly preferred in postverbal position in Spanish ‘why’-interrogatives with unergative verbs. Accordingly, focal subjects can be analyzed in the same way as in Italian. Non-focal postverbal subjects remain either in their merge position (e.g., Suñer 1994) or Spanish displays a further lower subject position that is not available in Italian where these subjects can move to, as proposed by Belletti (2004). In bare wh-interrogatives, postverbal subjects are assumed to remain in their base-position (Bocci and Cruschina 2018; Cardinaletti 2001). Importantly, postverbal subjects and final vocatives crucially differ with respect to their syntactic position. While postverbal subjects remain in the lower part of the syntactic tree, final vocatives occupy a position in the left periphery, as discussed in Section 7.1.

8 Conclusion

In this paper, we brought together two previously independent domains of linguistic research by investigating vocative structures in Spanish wh-interrogatives. To this end, we ran an acceptability judgment task to address the following research questions:

1. In which position are vocatives accepted in wh-interrogatives? (RQ1)

2. Do d-linked wh-phrases and por qué (‘why’) pose less restriction on the middle positions of vocatives than bare wh-elements? (RQ2)

3. Which position(s) do initial, middle, and final vocatives occupy in the left periphery? (RQ3)

Regarding RQ1, we found that initial and final vocatives are highly accepted in wh-interrogatives, regardless of the wh-phrase. This implies that initial and final vocatives are compatible with bare wh-elements, por qué (‘why’), and d-linked wh-phrases. For middle vocatives, the experimental results differed between wh-phrases. While the participants (at least marginally) accepted middle vocatives in wh-interrogatives introduced by por qué (‘why’) and d-linked wh-phrases, they rejected them in wh-interrogatives with bare wh-elements (RQ2). Put differently, d-linked wh-phrases and por qué (‘why’) pose less restriction on the middle positions of vocatives than bare wh-elements.

With respect to RQ3, we follow previous work and assume that the position vocatives occupy in the left periphery depends on their semantic-pragmatic function (González López 2019; Slocum 2016). Initial vocatives that have a call interpretation are located in Voc_callP, above ForceP (Espinal 2013; González López 2019; Hill 2013b; Moro 2003, a.o.). For middle and final vocatives that have the sematic-pragmatic function of an addressee, we proposed two Voc_addrPs in the left periphery. The upper one includes middle vocatives, and it is placed between the second TopP and the FocP. This extension of previous analyses explains the differences found between por qué (‘why’) and d-linked wh-phrases, on the one hand, and bare wh-elements, on the other. The lower Voc_addrP in which final vocatives are placed is located between the lowest TopP and the FinP. We assumed this second Voc_addrP for two reasons. First, it addresses the fact that Spanish allows for multiple addressees and second, we have seen that addressees can precede and follow focal constituents.

For future research, interesting follow-up questions would be whether other languages also reveal differences for middle vocatives with respect to wh-phrases and whether they allow for focal constituents preceding and following addressees.